Modelling the dynamics and control of stoats in New Zealand forests

The impact of sterilisation and culling control of stoat (Mustela erminea) populations was evaluated using models of increasing complexity. The first was a simple logistic model with continuous births and deaths; the second included a more realistic birth pulse rather than continuous births; and the third included a birth pulse and age structure. For beech forest (Nothofagus spp.) habitats, the birth pulse models distinguished between mast, crash and normal years, each year having a different intrinsic rate of increase, rm̳, which was parameterised from trap-catch indices. The second model best predicted the large variation in stoat abundance observed in beech forest. Using this model, little difference was predicted in the proportional reduction of stoat density under culling or sterilisation control. Under continuous control, sterilisation was slightly more effective at reducing peak (summer) stoat density; however under pulsed control, culling was marginally more effective than sterilisation. Control of either kind was much more effective against populations in non-beech forests than against those in beech forests, essentially because of the former population's lower rm̳ value. The second birth pulse model was also used to predict the likely dynamics of canine distemper virus (CDV) in stoat populations. CDV was not predicted to persist as an endemic disease in New Zealand stoat populations. This was primarily due to the birth pulse structure which precluded the continuous recruitment of susceptible individuals required to maintain the disease within a host population.